Throughout the explanation examples will be drawn from the
field of watershed management which is the application used
for implementing the concept.
4.1 Semantics of nodes
The metadata standards are structured in layers. Those parts
which are needed to uniquely identify nodes, contexts,
hierarchies, and classes, will be introduced in their respective
layer. Figure 2 shows a hierarchy of contexts classified into
nodes. Each node corresponds to a particular theme of
applications.
At node 0 the common syntactic vocabulary is defined, as
stated by OGIS. In addition to the common syntactic definition,
the semantics of nodes are also defined at this node. Nor
example node 1 is for earth resources and node 2 is related to
mapping. A node is a defined by the 7-tuple given by
sem <node> = <name, Address, sub-nodes, super-node,
context-thesaurus, role, contexts>
Name uniquely identifies nodes in the whole federation.
Address is the URL Internet address of the node.
Sub-nodes the immediate children of a node.
Super-node is the immediate parent of a node. Both super and
sub nodes will allow scanning the node/context hierarchy.
Context-thesaurus is a set of alternative names of
contexts within a node. This facility is added in order
to resolve the naming conflicts that can arise from
contexts, who have various perceptions and
understanding of another context.
Role is a list of data types provided by contexts which
belong to the underlying node.
having different remote users, i.e., from different |
As an example the semantics of node 1 and node 2 can
be as follows
sem (node 1) = < Earth resources, itc.nl, none,
federated GIS, (natural resources, environmental
data), (watershed data, water quality), (environment,
basin management)»
Node 2
sem (node 2) = « mapping, ma.nl, none, federated GIS,
(topography, surveying), {topographic information,
land parcel», «cadastral, maps, triangulation information»
4.2 Semantics Of contexts
Contexts refers to the assumptions underlying the way in which
an interoperating member represents or interprets data. Syntax
and semantics at node 1 are used as building blocks for
defining the immediate higher level contexts, i.e., WSM and
water quality in Figure 2. Hence syntactic and semantic
definitions are recursively defined from the lower level
contexts to the higher ones. The semantic relationships are
defined between each context and its super context. In this case
the semantic level for data exchange between two contexts is
the one defined at the common super context. For example the
semantic level for sharing data between analysis and
monitoring contexts is the regional context. A context is
defined by 14-tuple
sem (context) = <node, name, address, sub-contexts, super-
context, list of classes, class thesaurus, identification
information, distribution information, quality, role,
More Semantics
Base Map Heighet
62
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B3. Vienna 1996
association, spatial reference information, spatial data
organisation information>
Node the name of the node where the underlying context
belongs to.
Name is uniquely identifying the context in the whole
federation.
Address the URL Internet address of the context.
Sub-contexts are the immediate children of a context.
Super-context is the immediate parent of a context. Both super
and sub contexts will allow scanning the node/context
hierarchy.
List of classes is a set of class names which are in the database
of the underlying context.
Class thesaurus is the list of the alternative names of classes
within the underlying context.
Identification information basic information about the data
set as defined by FGDC. A Boolean value which indicates
whether the metadata exist or not.
Distribution information it is about the distribution options
for obtaining the data set, as defined by FGDC. A Boolean
value which indicates whether the metadata exist or not.
Quality is the general quality parameters of the database
contents as defined by FGDC. A Boolean value which
indicates whether the metadata exist or not.
Role is a set which represents the data types offered by the
underlying context.
Erosion Sediment Water quality
Environmental Social Economic
Monitoring Analysis Management
Local National
Regional
Hydrology Soil
Mapping Agriculture Earth Resources
Federated GIS
Node 0
Figure 2 Node And Context Hierarchy
Association A Boolean value which indicates if the underlying
class has a functional relationship with other classes.
Spatial reference information a Boolean value which
indicates whether there exist metadata.
spatial data organisation information is the mechanism used
to represent the spatial information in the data set. It is a
Boolean value which indicates whether there exist metadata
As an example consider the context of watershed management,
decision making for watersheds involves three types of
activities: 1) monitoring: where the aim is to prepare a concise
data inventory of the status of the watershed; 2) analysis:
where the aim is to analyse watersheds in order to assess their
vulnerability for degradation, and quantify the causes of
degradation; 3) management: where the intention is to
introduce and implement new management plans. These
activities are performed by different groups which can be
considered as different contexts. Corresponding to these
contexts are three object hierarchies Figure 3. At the lowest
level of these hierarchies are elementary objects. Elementary
objects are those at the lowest level of an abstraction hierarchy
of a particul
correspondi
managemen
further abst
and nationa
e At the
interpret
land de
subcatch
regional
e At the
abstracte
the sim
watershe
manager
Management
(MANAGEM
[rem
Soil
(
Figure 3 ,
polygoi
e At the
abstraci
provinc
Semantics
are define
sem (an:
analysis.it
felief), -T
manageme
Following
between
resources.
illustratioi
data type
and targ
degradatic
function i
